Macromolecules 1986,19, 3008-3010
3008
Spinodal Decomposition of Polymer Mixtures w i t h a Thermotropic Liquid Crystalline Polymer as One Component
10'
IOL
10"
io4
lo=
Figure 5. Distribution of decay times using various methods12 for data obtained on solutions of PS (M,= 15 X lo6) in cyclopentane (21"C) at a concentration of 4.7% (w/w) (measurements at 0 = Soo): (a)CON TIN^^ (continuousline); (b)linear-programming method," which uses fixed components equidistant on a logarithmic scale and minimizes the sum of absolute values of residuals (result shown as the broken line); (c) positive-coefficient exponential sum (PES)method,15which fits the experimentalcurve to the expression (CElcie-t/Ti)z+ b, with all ci, 7i,and b adjusted by least squares with ci > 0 and 7i > 0. Components N are increased until no new T~ with positive ci can be added. These values are illustrated as the vertical bars. The two vertical arrows shown are the results of bimodal analysis as described in ref 3. dynamics, not the least because it is difficult to obtain asymptotic conditions even with very high molecular weight fractions. We have endeavored to meet this need, as shown by the data in Figure 3. It is apparently necessary to exceed C N 10% (the often-used limit for semidilute solutions owing to monomer-friction complications) if a limiting exponent for the gel mode is to be established. However, the experimental difficulties are formidable in obtaining a precise description of the various components involved in the relaxation behavior. The degree of nonexponentiality is considerable, and substantial ambiguity attaches to the components regardless of the data-analysis technique employed. It is also possible that the slower modes couple to some extent with the faster motions of the transient gel a t the highest concentration^.^ Acknowledgment. The author is grateful to Dr. Petr SGptinek, Czechoslovak Academy of Sciences, for helpful discussions and for performing correlation function analysis. Registry No. PS,9003-53-6.
References and Notes Hecht, A.-M.;Bohidar, H. B.; Geissler, E. J. Phys. Lett. 1984, 45, L-121.
Adam, M.; Delsanti, M.Macromolecules 1985,18,1760. Brown, W.Macromolecules 1986,19,387. Brochard, F.;de Gennes, P.-G. Macromolecules 1977,10,1157. Brochard, F. J. Phys. (Les Ulis, Fr.) 1983,44, 39. de Gennes, P.-G. Scaling Concepts in Polymer Physics; Corne11 University Press: London, 1979. Chu, B., personal communication. Brown, W.; Johnsen, R. M. Macromolecules 1985, 18,379. Noda, I.; Higo, Y.; Ueno, N.; Fujimoto, T. Macromolecules 1984, 17, 1055. These authors show, using osmotic pressure measurements, that the crossover to the concentrated region occurs at between 15 and 20%, irrespective of molecular
weight.
Brown. W.Macromolecules 1986., 18. , 1713. ALV-Langen co., F.R,G. Sthiinek,P.;Kofiilk, C.; Jakei, J. Polym. Bull. (Berlin) 1986, 16,67. Provencher, S. W.Makromol. Chem. 1979,180,201. Zimmerman, K.; Delaye, M.; Licinio, P. J. Chem. Phys. 1985, 82,2228. Jakeii, J.; SGpinek, P., to be published.
Wyn Brown Institute of Physical Chemistry University of Uppsala 751 21 Uppsala, Sweden Received June 18,1986 0024-9297/86/2219-3008$01.50/0
The mechanism and kinetics of phase separation of polymer mixtures have been the subject of extensive studies in recent y e a r ~ . ' - ' ~ ~ The ~ " ~studies ~ involve timeresolved analyses of structural evolution from the mixtures in the molten, isotropic liquid state1-8J1J3,35-39 or in a neutrally good solvent9J0J2 (i.e., phase separation into polymer A solution and polymer B solution), induced by quenching the systems to a thermodynamically metastable or unstable state. In this paper we present novel experimental evidence on phase separation into a liquid crystalline phase and a molten, isotropic, liquid phase via spinodal decomposiDuring the course of the isothermal phase separation we observed the following: (i) We found an evolution of periodic concentration fluctuations with correlation length A, which increases with time spent for phase separation. (ii) The time evolution of the concentration fluctuations results in a bicontinuous, periodic networktype structure: one type of phase is rich in liquid crystal forming polymer and forms the liquid crystalline phase, and the other type of phase is rich in molten, isotropic polymers and forms the optically isotropic phase. (iii) At a sufficiently late stage, a "~elf-similarity"'~~~~ was found between the structures obtained at different times. Observations i-iii are essentially the same as those reported for unmixing processes of critical mixtures of metallic alloys,1e20 small-molecule liquids,21v22 inorganic gla~ses,2~ and binary liquid polymers.24,40It is our objective in this paper to report that this common phenomenon can also be found in the unmixing process of binary mixtures between thermotropic, liquid crystalline polymers and isotropic liquid polymers. The system used in this study was a 5050 mixture (by weight) of a commercial poly(ethy1ene terephthalate) (PET) and a copolyester of 60 mol % p-oxybenzoate (OBA) units and 40 mol % ethylene-terephthalate units (X-7G; Tennessee Eastman Co.). The number-average molecular weights of these polymers are both ca. 2 X lo4. X-7Gis among the first thermotropic liquid crystalline polymers, and its synthesis and properties have been reported by Jackson and K u h f u s ~ .These ~ ~ two polymers were dissolved in o-chlorophenol at 60 "C, and the insoluble part, which probably consists of crystalline poly(poxybenzoate), was removed by filtration. The test specimens were prepared by casting one or two droplets of 1-5 wt % solutions on microscope cover glasses. The solvent was quickly evaporated by placing the cover glasses onto a hot plate preheated at 60 "C in a vacuum chamber. The film-casting seemed to be completed within a few minutes but the test specimens were kept under vacuum for an additional 3 h. The films thus prepared had average thicknesses of 1-10 pm at the centers of the droplets and thicker peripheries. The as-cast films were clear and exhibited no optical anisotropy under a cross-polarized optical microscope. Time evolution of the structure was observed by a Nikon Optiphot-Pol X T P - 1 1 polarizing microscope equipped with a TH-600 heating stage (LINKAM Scientific Co.). The temperature of the heating stage was calibrated for the melting points of various standards of small molecules. The observations were started as soon as the test specimens on the cover glasses were placed on the heating stage controlled at the preset temperature. The time required for a temperature jump (2' jump), i.e., the time necessary for the test specimens to reach the preset temperature, was less than a few seconds. The 0 1986 American Chemical Society
Communications to the Editor 3009
Macromolecules, Vol. 19, No. 12, 1986
X - 7 G / P E T ( 5 0 / 5 0 ) at 256°C X-7G/PET ( 5 0 / 5 0 )
101 os cast
lbl 4 set.
P
-!?El
IcI I 2 sec
280°C
- R.T.
258 " C
- R.T.
Figure 2. Polarized optical micrographs of X-'IG/PET (50/50) mixture film: slowly cooled to room temperature after phase separation at 280 "C (left) and rapidly eooled to rcom temperature after phase separation at 258 "C (right). Id1 2 6 sec.
le1 3 8 s e c .
X-7G/PET (50/501 at 238°C 8sec.
Figure 1. Polarized optical micrographs of X-lG/PET (50150) mixture film: (a) observed at room temperature (as-cast film) and observed at 256 "C after (b) 4, (c) 12, (d) 26, and (e) 38 s of isothermal phase separation at 256 "C. polarized optical micrographs were taken with the increment of annealing time. The light scattering patterns were obtained, however, from specimens quenched after various annealing times on a metal block a t room temperature. Figure 1shows the structure evolved by the isothermal phase separation of the mixture a t 256 'C. The isothermal phase separation was initiated by a T jump from room temperature t o 256 O C for the as-cast films. The as-cast films are o p t i d l y isotropic and homogeneous on a spatial scale of 50 wm as shown in Figure la. The temperature 256 "C is above the melting point of the crystallites of the P E T sequence in X-7G and pure PET" so that PET homopolymer and the P E T sequence in X-7G are in the molten liquid state. X-7G forms liquid crystals at 256 "C.= Shortly after the T jump, unmixing between the optically isotropic phase and the anisotropic phase was observed to occur, as shown in the pictures obtained 4 and 12 s after the T jump (Figure lb,c, respectively). The phase separation appears to form periodic, bicontinuous network structure, and the periodicity or the correlation length appears to increase with time. With a further lapse of time, a coarsening of the network pattern occurs, while "self-similarity" of the patterns seems to be maintained throughout the process. The fine structures observed within the optically anisotropic networks are believed to originate from a polydomain structurez6 in the thermotropic liquid crystal phase. The polydomain structure in turn is believed to arise from the spatial distribution of the disclination lines2' and hence to represent regions of the correlated orientations26 of the OBA optical axes. Except for the novelty associated with one phase being liquid crystalline, the general trends observed in the time evolution of the unmixing structures are typical of those for unstable critical mixtures in small-molecule liquids, metallic alloys, inorganic glasses, and binary liquid polymers. Figure 2 demonstrates that the isotropic phase in Figure 1 indeed corresponds to a liquid PET phase. The left picture was obtained by slowly cooling the mixture subjected to a certain stage of unmixing a t 280 "C to room temperature, while the right picture was obtained by rapidly cooling the mixture subjected to another stage of unmixing a t 258 OC to room temperature. In the left picture, the defocused regions located in the left and right corners of the pictures correspond to a part of the liquid
H"
V"
HH
Figure 3. Light scattering patterns taken under Hv ( I C , L I , Vy (center),and H, (right) conditions from an X-lG/PET (G0/50) mixture film isothermally phase separated at 238 "C for 8 s and quenched to r w m temperature. The directions of the polarizer and analyzer are shown by arrows. crystalline networks which are frozen in the glassy state. Between the liquid crystalline regions one can clearly observe the PET spherulites that grow from the isotropic liquid phase during the cooling process. In the right picture, a part of the glassy liquid crystalline networks is observed in the lower part of the picture, and small PET spherulites or crystallites are observed in the matrix. The time evolution of the light scattering patterns during the isothermal phase separation was also investiDuring phase separation ring-like scattering patterns typical of what is called a 'spinodal-ring"g was observed, and the peak position was found to shift toward the incident beam axis. The time evolution of the ring-like scattering reflects growth of periodic concentration fluctuations originating from the spinodal decomposition and its coarsening process. The light scattering technique was found to be especially powerful in the investigation of the early stage of unmixing where the periodic network structure cannot be clearly resolved under the optical microscope due to overlap of the structure in the thickness direction. Of course, good agreement between the correlation length Am's measured from the light scattering technique and optical microscopy was clearly found in the late stage of unmixing where the network structure is clearly distinguished under the microscope. Figure 3 shows a unique feature of the spinodal ring that results from one phase being optically anisotropic due to the liquid crystal nature. The anisotropy gives rise to the strong depolarized (Hv) pattern. The orientations of the polarized and depolarized patterns are expected from scattering theory for anisotropic systems26s3z and indicate preferential orientation of the OBA molecules in the liquid
Macromolecules 1986, 29,3010-3011
3010
c
1
t
t
I
1
I
I
10
100
T i m e / sec.
Figure 4. Logarithmic plot of the correlation length A m vs. time spent for the isothermal phme separation of X-7G/PET (50/50) mixture film a t 256 "C.
crystalline networks. Detailed analyses on the scattering patterns will be described elsewhere.2s Figure 4 shows the time evolution of the correlation length Am's during isothermal phase separation at 256 OC. Am's were measured by image processing of the optical microscopic pictures obtained in the late stage of the unmixing so that the network structure could be unequivocally distinguished with the dynamic image analyzer built in our l a b ~ r a t o r y .Details ~~ of the dynamic image analyzeP and the method of extracting Am's will be reported elsewhere.28 Preliminary experimental results show the power law Am t1/3 (1) The '/,-power law is the one familiar for the droplet growth caused by the hydrodynamic e f f e ~ t ~ ~or J ' 9by~ the Lifshitz-Slyozov process.16 The growth of the droplets occurs as a consequence of their diffusion and coalescence so that
-
Rqt)
- D(t)t
(2) where R ( t ) is the size of the droplet and D ( t ) is time-dependent diffusivity of the droplet, which, in turn, is given by D ( t ) kBT/VR(t) (3)
-
where kBT is the thermal energy and q is the viscosity. From eq 2 and 3 one obtains the '/,-power law
- -
R ( t ) (kBT/?7)'/3t'i3 (4) Needless to say, further investigation is required as to whether the '/,-power law observed in the preliminary study of the unmixing dynamics of the liquid crystalline component and isotropic liquid component exactly corresponds to the '/,-power law in the growth of patterns caused by hydrodynamic interactions of the LifshitzSlyozov process found for the isotropic liquid or solid mixtures. Moreover, the 1/3-power law itself has to be further investigated over a much wider time scale and temperature range. Further work along this line is currently in progress in our laboratory. However, all evidence presented in this paper indicates that spinodal decomposition and subsequent coarsening processes similar to those for isotropic liquids and solids occur even when one component is a liquid crystalline polymer and forms a liquid crystal phase. Acknowledgment. We are grateful to Dr. S. Suehiro for help in the image processing of the optical micrographs with the dynamic image analyzer. Part of this work was supported by the Japanese Ministry of International Trade and Industry (MITI). 0024-9297/86/2219-3010$01.50/0
References and Notes Nishi, T.; Wang, T. T.; Kwei, T. K. Macromolecules 1975,8, 227. de Gennes, P.-G. J . Chem. Phys. 1980, 72, 4756. Nojima, S.; Tsutsumi, K.; Nose, T. Polym. J. (Tokyo) 1982,14, 225. Nojima, S.; Ohyama, Y.; Yamaguchi, M.; Nose, T. Polym. J . (Tokyo) 1982,14,907. Hashimoto, T.; Kumaki, J.; Kawai, H. Macromolecules 1983, 16, 641. Snyder, H. L.; Meakin, P.; Reich, S. Macromolecules 1983,16, 757. Binder, K. J. Chem. Phys. 1983, 79, 6387. Gelles, R.; Frank, C. W. Macromolecules 1983, 16, 1448. Hashimoto, T.; Sasaki, K.; Kawai, H. Macromolecules 1984,17, 2812. Sasaki, K.; Hashimoto, T. Macromolecules 1984, 17, 2818. Izumitani, T.; Hashimoto, T. J. Chem. Phys. 1985,83, 3694. Inoue. T.: Oueisawa. T.: Yasuda. 0.: Mivasaka. K. Macromolecules' 1 9 6 , 18, 57, ' Russel, T. P.: Hadziioannou, G.; Warburton, W. K. Macromolecules 1985, 18, 78. Cahn, J. W. J . Chem. Phys, 1965,42,93. Lifshitz, I. M.; Slyozov, V. V. J. Phys. Chem. Solids 1961,19, 35. Binder, K.; Stauffer, D. Phys. Reo. Lett. 1974, 33, 1006. Furukawa, H. Prog. Theor. Phys. 1978,59, 1072; Phys. Reu. Lett. 1979,43, 136; Phys. Reu. A 1981, 23, 1535; Physica A: (Amsterdam) 1984, 123,497. Hennion, M.; Ronzaud, D.; Guyot, P. Acta Metall. 1982, 30, 599. Komura, S.; Osamura, K.; Fuji, H.; Takeda, T. Phys. Reu. B 1984,30, 2944. Furusaka, S.; Ishikawa, Y.; Mera, M. Phys. Reu. Lett. 1985,54, I
I
~
2611.
thou, Y. C.; Goldburg, W. I. Phys. Reu. A
1981, 23, 858. Knobler, C. M.; Wong, N. C. J . Phys. Chem. 1981, 85, 1972. Craievich, A.; Sanchez, J. M. Phys. Reu. Lett. 1981, 47, 1308. Hashimoto, T.; Itakura, M.; Hasegawa, H. Polym. Prepr. Jpn., SOC.Polym. Sci., Jpn. 1985,34, 2773; J . Chem. Phys., in press. Hashimoto, T.; Itakura, M.; Shimidzu, N. J . Chem. Phys., in press. Jackson, W. J.; Kuhfuss, H. F. J . Polym. Sci., Polym. Chem. Ed. 1976, 14, 2043. Hashimoto, T. J. Appl. Polym. Symp. 1985, 41, 83. Demus. D.: Richter. L. Textures of, Liauid . Crvstals: Verlae Chemie: New York, 1978. Nakai, A.; Shiwaku, T.; Hasegawa, H.; Hashimoto, T., to be submitted. Stein, R. S.; Rhodes, M. B. J. Appl. Phys. 1960, 31, 1873. Rhodes, M. B.; Stein, R. S. J . Appl. Phys. 1968, 39, 4903. Hashimoto, T.; Ijitsu, T.; Yamaguchi, K.; Kawai, H. Polym. J . (Tokyo) 1980.12. 745. Haskmoto, T.; Ebisu, S.; Inaba, N.; Kawai, H. Polym. J . (Tokvo) 1981, 13. 701. Suehiro, S.; Takebe, T.; Izumitani, T.; Hashimoto, T., to be submitted. Kawasaki, K.; Ohta, T. Prog. Theor. Phys. 1978,59, 362. McMaster, L. P. Adu. Chem. Ser. 1975, No. 142, 43. Gilmer, J.; Goldstein, N.; Stein, R. S. J. Polym. Sci., Polym. Phys. Ed. 1982,20, 2219. Hill, R. G.; Tomlins, P. E.; Higgins, J. S. Macromolecules 1985, 18, 255. Han, C. C.; Okada, M.; Muroga, Y.; McCrackin, F. L.; Bauer, B. J.; Tran-Cong, Q. Polym. Eng. Sci. 1986, 26, 3. Yang, H.; Shibayama, M.; Stein, R. S.; Shimidzu, N.; Hashimoto, T. Macromolecules 1986, 19, 1667. Takahashi, M.; Horiuchi, H.; Kinoshita, S.; Ohyama, Y.; Nose, T. J . Phys. SOC.Jpn. 1986,55, 2687. Wide-angle X-ray diffraction and small-angle X-ray scattering showed the melting point of the pure P E T is close to 275 OC, slightly higher than 256 O C . However, the crystallinity of the P E T homopolymer in the mixture, if it existed, appears to make an insignificant contribution to the anisotropy of the PET-rich phase.
I
Akemi Nakai, Toshio Shiwaku, Hirokazu Hasegawa, and Takeji Hashimoto* Department of Polymer Chemistry Faculty of Engineering Kyoto University, Kyoto 606, Japan Receiued July 25, 1986
0 1986 American Chemical Society
